Warmly, the autonomous sales orchestration platform, went from zero to 100+ paying customers in 2023. More importantly, we went from founder-led sales to a repeatable sales process.
We recently raised an $11 million Series A led by Felicis, helping us continue the momentum into 2024.
Introduction to Warmly's Success Story
It wasn’t all up and to the right even if it might look that way from the outside (founders getting awarded Forbes 30 under 30, participating in both Y Combinator and Techstars).
Warmly was founded in 2019 (yes, you heard that right) and our initial product was a virtual name tag for Zoom. While the original product won accolades (like this PLG123 video) and was a Zoom Apps launch partner, ultimately we struggled to sell it. We had a second major product pivot as well before focusing on the current iteration during the second half of 2022.
Because we already had another product in-market, we built the sales orchestration platform in stealth for six months before sharing it publicly. This new product gets customers warm leads by de-anonymizing website traffic and then automatically following up with visitors via chat, video, email, or LinkedIn. This helps customers generate 10x more outbound leads than a traditional SDR while seeing 3x higher close rates because these are warm leads rather than cold ones.
Keep reading for our color commentary on how we built a repeatable go-to-market motion in 12 months and what we learned along the way.
Q1 2023: Founder-led sales
- 👨👩👧👦 Customers: 0 → 10
- ⚡️ GTM strategy: Design partners and founder friends
- 📈 Key experiment: Messaging
- ❓ Key questions:
- (1) Can we sell Warmly at any price point?
- (2) Can I stand up a basic funnel?
- (3) Can I find the 10 most common objections and solve them?
- 🛠️ Key new tools: HubSpot (CRM), Mixmax (sales execution), ConnectTheDots (warm intros), DocuSign, spreadsheets
Entering Q1, “everything but sales were already in place,” and so I could spend the majority of his time scaling go-to-market.
Our main initial experiment was on Warmly’s messaging. I needed the words for a sales person to be able to talk about what we were doing. Since we had previously built a solution that was different from the problems that customers were trying to solve, I was particularly keen to avoid making that mistake again.
I began by looking at SEO rankings and researching what others in the space called themselves. One tactic is to go to every competitors’ website, write down all of the words on their homepages, and then build a word cloud around it. “Competitors spend five to six years figuring out how messaging works so why reinvent the wheel?”
One of my early email outreach messages
There were a bunch of category and message variations in Warmly’s market (ex: account-based marketing, contact databases, intent data). I tested some of these messages with potential design partners, allowing him to find out what competitors got wrong and what we got right. I then tested a message focused on a live video caller Slackbot that would capture people before we leave the website. That failed to land – people instead wanted outreach to be done for them in an automated way and done across multiple channels rather than just video chat.
We landed on autonomous sales orchestration, which captured what customers were trying to do and was meaningfully different from what competitors talked about. The simplified message was that Warmly helped you “get warm leads and talk to them live.” We could tell that it was working because prospects were nodding their head along and would say “here’s how I see that being used here”. (As an aside, I thought about AI sales orchestration, but found that people were sick of AI as a marketing message.)
I knew that Warmly's messaging wouldn't be completely solved in only a quarter. We were able to specifically tackle (a) what kinds of other tools did people have?, (b) what’s the category that we operate in?, (c) what phrases will get people to join a demo?, and (d) what phrases were other people using?
Another priority for Q1 was to set up a basic sales funnel.
Can I see that if I talk to roughly 10 people, one will close? If I talk to 20 people, will two close?
I had heard general benchmarks about having 5x coverage; that is, for every five qualified conversations, you close one deal. As a founder, I didn’t worry about hitting these exact benchmarks early on because he knew he was talking to “the most random people who weren’t in our ICP” and therefore would never buy. My bigger concern was the close rate among qualified prospects and finding out who was in Warmly’s ICP.
As a product with a $10-15k average annual contract value (ACV), I was looking to see that these deals could be closed in only a few meetings and with a 30-45 day sales cycle. Seeing positive signals, he was ready to scale the magic.
A snapshot of Warmly’s funnel stages for their outbound pipeline (we have separate pipelines for freemium, midmarket, and inbound)
Q2 2023: Sales leader-led sales, founder involved
- 👨👩👧👦 Customers: 10 → 30
- ⚡️ GTM strategy: Founder friends, startups, email sequencing
- 📈 Key experiment: LinkedIn sequencing
- ❓ Key questions:
- (1) Can a strategic sales leader who is a non-founder roughly hit projected quota?
- (2) Can I as the founder hit quota and also reasonable terms (no opt-outs, annuals only, discounts)?
- (3) Can I build trust with this leader to sell my vision and build a team around them?
- 🛠️ Key new tools: Warmly/6sense (website de-anonymization), Outreach (sales execution), Seamless.ai (contact database, Warmly (AI website chatbot), Sendspark (video personalization), Warmly/Salesflow.io (auto-LinkedIn sequences), LinkedElf (LinkedIn connections)
After starting to prove out founder-led selling, I chose to recruit a sales leader and then hire sellers after that. This would've extended our go-to-market progress by a quarter if not for the fact that our early sales hires used to work at the same company as this new sales leader. If we had to recruit, ramp up, and train new sellers – particularly sellers who might not have been effective – it would’ve taken much longer.
I emphasized that I wasn’t looking for a sales leader from a really big, later stage company. It was important to me that the sales leader be scrappy, hungry, entrepreneurial, and experienced – but not someone who’s done it twenty times. If I couldn’t find someone like that, I believed that the safer path for Warmly would’ve been:
- Founder-led sales first
- Then hire a BDR/SDR to scale the founder
- Bring on the first AE to report to me
- Once that’s successful, hire a second AE
- Then, hire a sales leader
My big experiment for the quarter was LinkedIn sequencing as a source of qualified pipeline. I had heard from other go-to-market leaders that email wasn’t working as well as it used to and that LinkedIn sequencing was still effective.
My co-founder Alan, Warmly’s SDR leader, and I adopted LinkedElf and Salesflow to max out our LinkedIn connections, adding 100 sales and marketing leader connections every week on auto-pilot. Then we wanted to do messaging on auto-pilot, too. We were able to send out 300 connection requests a week, seeing about half of them accept, and then one-eighth of them reply (half of which would be positive replies, the other half not so much…). Collectively, we were able to book about 10-15 meetings per week just from conversations on LinkedIn – and found a channel that we could continue to grow Warmly’s pool of leads.
A refined version of Warmly’s messaging now that we have more customer validation
Q3 2023: Seller-led sales, founder involved
- 👨👩👧👦 Customers: 30 → 60
- ⚡️ GTM strategy: Omni-channel (Linkedin/email), inbound (warm calling on the website with Warmly), territories
- 📈 Key experiment: Conferences
- ❓ Key questions:
- (1) Can an AE and ISR ramp and hit quota in the last month of the quarter?
- (2) Can our sales leader enable the sellers to ramp and hit quota?
- (3) Can we build a V1 sales process that can be understood and executed?
- 🛠️ Key new tools: Warmly (AI auto-emails), Warmly (AI auto-LinkedIn), HubSpot Quotes ,Tourial (pre-made demos), Spekit (sales process documentation), AccountAim (territory building)
At this point, our new sales leader and I needed to quickly build out the go-to-market team. First, we made a financial model to check whether the math would work out. If you’re not reconciling your quota against your financial model, you’re screwed. The financial model allowed our team to see things like:
- When is the next milestone we need to fit for fundraising?
- How do we get there without being cash-out?
- If Warmly had these milestones, how many reps do we need and when do we need to hire them?
- What would quota have to be in order for all of this to work out?
While in the past mature SaaS companies might’ve aimed for a 5x quota to on-target earnings (OTE) ratio, I didn’t believe that was realistic given the current buying environment and given that our sellers would need to source some of their own leads. We instead aimed for a 3x quota:OTE ratio and a 50/50 split between base and OTE commission, which would fit Warmly’s financial model and allow us to attract the right caliber seller. (As an aside, I encourage others to pay a higher salary-to-commission ratio in the early days so you don’t lose talent while you’re figuring out what’s working and what’s not working.)
Here’s what the team structure ultimately looked like:
- Sales leader (KPI = overall revenue that was sales-sourced / sales-closed)
- Account executives, i.e. AEs (KPI = closed-won revenue)
- Inbound sales reps, i.e. ISRs (KPI = closed-won revenue for very small deals and qualified opportunities among inbound demo requests)
- At Warmly, the ISR role creates a path for SDRs to get promoted; ISRs can advance to AEs if they consistently hit quota
- SDRs (KPI = sales qualified opportunities)
- We do international SDR hiring, which helps keep costs down, and has a 1:1 ratio between AEs and SDRs
- Sales assistants, aka SAs
- The SA helps with sales admin work, joins demo calls, drafts replies for AEs, pulls lead lists, and writes internal and external follow-up notes
- We got our sales assistants via Virtualis
In Q3, we bet on attending conferences to diversify pipeline generation beyond LinkedIn, greenlighting five conferences for the quarter. For each conference, our sales leader Keegan Otter would attend with one sales rep. We decided not to buy a booth for any of them; rather, we simply went, were friendly, and tried to meet as many attendees as we could.
Three of the five events turned out to be successes; two were not, including Dreamforce which was both extremely large and lacked buyers in our ICP. Overall, the conferences generated more money in closed won sales than it cost for us to attend. But there were drawbacks. For example, I found that there were tons of meetings booked from conferences, but also a spike in demo no-shows. And it was draining for the team to be on the road for five events over the course of only six weeks.
Reflecting on the quarter, I believe that most companies find three to four channels that work for them. It’s important to try all the channels early on, then get laser focused and double down on the ones that work. Conferences were a ‘tweener’ – we needed to keep iterating.
Q4 2023: Seller-led and sales leader run
- 👨👩👧👦 Customers: 60 → 100+
- ⚡️ GTM strategy: No change from Q3 (repeatability!), just optimization
- 📈 Key experiments: LinkedIn social to drive inbound, PLG
- ❓ Key questions:
- (1) Can I as the founder step away from closing altogether?
- (2) Can the sellers hit quota each month across the quarter?
- (3) Can we find repeatability in our core metrics (meetings held, SQOs, closed/won) and rates?
- 🛠️ Key new tools: Letterdrop (LinkedIn social), Alysio (gamified daily sales metrics), Warmly (cross-tool autonomous sales orchestration)
By Q4, Warmly’s go-to-market started to look more and more repeatable. My focus turned to optimization.
We bet on product-led growth (PLG), which Warmly had gotten good at during a previous pivot and hadn’t yet applied to the current product. Unlike the Zoom name tag product, the sales orchestration platform involves extra friction for self-service adoption – users have to install a script on their website.
Warmly’s pricing page includes a free account for smaller customers
We got creative with the free offering, designing it to be an on-ramp to Warmly’s core product rather than a replacement. Free users would get access to up to 500 free warm leads each month along with intent signals and alerts about those leads. If users then want automation to act on those leads, that’s when they'd have to pay for it. There have now been over 1,000 installs of our PLG offering and we're still early in optimizing free-to-paid conversion.
Two key learnings that made PLG different this time around:
- PLG became an acquisition strategy for a more valuable product. Before the pivot, we built a great product but people weren’t willing to pay for it. This time we validated that we could sell it before investing in PLG.
- PLG helps segment Warmly’s prospects. We thought if we had both ‘book a demo’ and ‘try for free’ on the website, everyone would click ‘try for free’. It turns out that companies with less than 50 people naturally sign up for free while companies with more than 50 request a demo. That’s what we wanted.
Reflecting on the last 12 months – and what comes next
First tip: your CRM setup is critical (we chose HubSpot). At first I was very anti having a bunch of form-fills and I understand why reps hate updating a CRM, but the better you can do this in the early stage, the better. I would recommend five fields to always fill out:
- How did we hear about us?
- How was the meeting booked? (Get really specific – Warmly has 20 sub-channels – because that will tell you what channel to double down on.)
- If you closed-won or closed-lost to a competitor, which one?
- Who was the associated AE?
- Who was the associated SDR? (An associated SDR might be the same person as the AE if your AE’s are full-cycle.)
The beauty of CRM reporting when you have good data (the actual numbers have been anonymized)
Second: clean deals are better. The sooner you can get away from month-to-month, opt-out, weird deal terms, do it. It becomes a big headache for CSMs and for managing how you pay out reps.
Third: pay out your sales leader on a percentage of overall revenue that was sales-sourced and sales-closed (i.e. everything except for deals that are founder-sourced and closed). I don’t give commission out on deals that I find and close on my own as a founder. But any deal where an AE is involved or an SDR is involved, the sales leader gets a commission on overall revenue to align incentives.
Fourth, and finally: you and your sales leader should stay accountable for closing deals on your own. I still try to self-source and close three deals a month and my sales leader does, too. You have to be so close to the process and find issues with it before taking a step back.